Open-Load Protection for Wireless Power Transfer Systems

[Inna89K]

What are the common methods for protecting a wireless power transfer (WPT) system with 0.2 coupling coefficient and without a feedback link against an open-load condition? In our design, when the load is disconnected, the transferred power is dissipated almost entirely in the inductors of the receiver-side matching network. As a result, these inductors overheat and can fail. What protection techniques are commonly used to prevent this? 

[mtwieg]

I believe most systems rely on communications from the receiver back to the transmitter (either via backscatter modulation or some other channel) to handle scenarios like this.

Without some sort of feedback, how is the transmitter supposed to tell the difference between power being absorbed by the receive coils and power absorbed by an actual load on the receiver?

[jbb]

Some quick ideas:
- could you de-tune the receiver a bit (yes, this would sacrifice some performance/efficiency) so that it can tolerate the fault condition?
- would the primary controller see enough difference between an OK receiver and an about-to-blow-up receiver to do protection (By measurement at primary side, not communications)?
- could you place a clamp system closer to the receiver coils before any tuning capacitors come into play? That way you’d be shunting the L of a pickup coil instead of the LC (and very small R) or a tuned pickup
- is there a role for saturating inductors in there somewhere?
- is a (switched?) dummy resistor on the receiver output a possibility?

[moffy]

What are the common methods for protecting a wireless power transfer (WPT) system with 0.2 coupling coefficient and without a feedback link against an open-load condition? In our design, when the load is disconnected, the transferred power is dissipated almost entirely in the inductors of the receiver-side matching network. As a result, these inductors overheat and can fail. What protection techniques are commonly used to prevent this?

If there is no communication then the protection would best be on the receive side, over current monitoring/temp/protection.

[PCB.Wiz]

What are the common methods for protecting a wireless power transfer (WPT) system with 0.2 coupling coefficient and without a feedback link against an open-load condition? In our design, when the load is disconnected, the transferred power is dissipated almost entirely in the inductors of the receiver-side matching network. As a result, these inductors overheat and can fail. What protection techniques are commonly used to prevent this?

The removal of a load usually also removes damping from resonant circuits, meaning the higher voltages and currents that you are worried about, can be sensed.
You basically measure the Q on a continual basis.

[Inna89K]

- I'd rather avoid detuning the system, since efficiency is one of the key requirements. Also, because the coupling between the coils is strong, the receiver coil will still see a high induced voltage, even with a detuned resonant circuit.
- There is no controller in the transmitter.
- A protection circuit could be placed on the receiver coil PCB, but it would have to operate in a strong magnetic field.
- Yes, the inductors are used to filter the switching noise generated by the rectifier diodes.The required inductance value, together with the available size constraints, makes the use of air-core inductors impractical.
- No, there isn't enough space for that.

[mtwieg]

- I'd rather avoid detuning the system, since efficiency is one of the key requirements.

Obviously efficiency isn't relevant in the no-load case. Just need to make sure the "detuning" function doesn't significantly degrade efficiency during high load.

Also, because the coupling between the coils is strong, the receiver coil will still see a high induced voltage, even with a detuned resonant circuit.

I don't think this should be an issue, unless you're using an unconventional rectifier circuit which connects directly to the RX coil. Can you show a basic schematic?

- There is no controller in the transmitter.

Is it just constantly switching at a fixed frequency? Or does it automatically lock to the coil resonant frequency via feedback?

[Inna89K]

The issue was not caused by an open load, but by a short circuit across the smoothing capacitor at the rectifier output. A batch of PCBs contained defective capacitors that caused the output to short, resulting in excessive power dissipation in the Lf inductors, which eventually burned out.

An open-load condition does not lead to this problem because the system operates in a different mode, where the transmitter delivers only a small amount of power.

Nevertheless, it would probably be worth adding protection against this failure mode as well?

[moffy]

Let me guess, MLCC capacitor? You might want to put a fuse in series with it.

[mtwieg]

Nevertheless, it would probably be worth adding protection against this failure mode as well?

If the overheating could potentially cause a fire/safety hazard, then yes.

Let me guess, MLCC capacitor? You might want to put a fuse in series with it.

Probably should be in series with the rectifier output rather than the capacitor itself (guessing this is what you meant).

[Inna89K]

Yes, MLCC. Is this a common issue?

[moffy]

Yes, MLCC. Is this a common issue?

Yes, exceedingly common. It appears to be the cause of a lot of laptop failures.

P.S. I suspect that some of the ceramics of MLCC are more piezoelectric than others and might suffer fatigue, but that's just a suspicion. Some more knowledgeable person might be able to clarify that.